Although health care spending in the U.S. will likely increase by 25% by 2030 due to lack of understanding of how to preserve older adults' health (Centers for Disease Control and Prevention and The Merck Company Foundation, 2007), ethnically diverse older women are a highly neglected research population. They are particularly ignored in research on the links among trauma exposure, PTSD, and health. Limited knowledge of these clinical issues could be contributing to the high utilization and related cost of health care for this mainly economically disadvantaged population, as older women with unaddressed PTSD symptoms often have unexplained somatic symptomatology, which leads to consulting doctors frequently and with discouraging results (van Zelst et al., 2006). Thus, investigating their trauma exposure, PTSD, and related health has the potential to save money, energy, and visits to doctors, among many other benefits. To fill several gaps in the ethnogeriatric literature, this study's aims are: (1) to assess the prevalence of PTSD and investigate the relationship between lifetime traumatic experiences and PTSD symptomatology in a non-clinical sample of multiethnic older women; (2) to validate a new post-traumatic stress screen for use in fast-paced medical settings using a sample of ethnically diverse older women, and (3) to test a model examining the relationship between lifetime traumatic experiences and health in community-dwelling older women accounting for a variety of potential protective factors. To this end, the research team will recruit approximately 230 community-dwelling women age 65 or older from diverse ethnic backgrounds. In a manner that will minimize fatigue and attrition, we will assess respondents using a clinical battery that includes multiple measures of trauma exposure, PTSD symptomatology, psychosocial and cultural resources, as well as health (physical and mental). The PTSD and health models proposed in Aims 1 and 3 are based on relevant theoretical and empirical literature; testing of such models will occur through application of structural equation modeling procedures. The main hypotheses are that the 2 models are valid, thus explaining a significant amount of variance in PTSD and health, respectively, and that our screen is a reliable and valid tool, upon validating it against the full-length PTSD measure from which it was derived, the DEQ (Kubany et al., 2000) and the gold standard of PTSD assessment, the CAPS (Blake et al., 1990). Our main long-term goal is to make a strong contribution to the development of health risk profiles and interventions that could positively impact health in our understudied population. )